Kazutaka Maruyama

Discovery of Interesting Users in Twitter by Overlapping Propagation Paths of Retweets

In recent years, social networking services have come into wide use to people. Especially, one of micro blog services, Twitter is a significant service. Twitter user gets information by following other users whose tweets match his interest. Retweet is one of Twitter functions which spreads tweets to other users. Using retweets, one can read tweets originated by users who are not followed by him. Our goal is to discover Twitter users who retweet many tweets which match the interest. We focus on the propagation of retweets and build a graph, the Overlap Graph, which contains users who share same retweets. Finally, we validate the users appearing in the graph by checking the frequency and the content of their retweets.

Japanese vowel recognition using external structure of speech

Speech acoustics is inevitably distorted by non-linguistic features such as vocal tract length, gender, age, microphone, room, line, hearing characteristics, and so on. Recently, a novel acoustic representation of speech was proposed, called the acoustic universal structure [1, 2]. It discards all the absolute properties of speech events and captures only their interrelations or contrasts to represent external structure of speech. Based on a mathematical model of the non-linguistic distortions, the new representation can remove the inevitable distortions as cepstrum smoothing can remove pitch information from speech. Recognition experiments using the external structure were conducted and it was shown that the proposed structure models trained from a single speaker with no normalization can outperform the conventional speaker-independent HMMs with CMN and SS. It was also found that lowpass filtering and white noise addition as preprocessing improved the performance because they can suppress the distortions rather well

Development of a Femininity Estimator using Speaker Recognition Techniques for Voice Therapy of Gender Identity Disorder Clients

This paper describes the development of an estimator of perceptual femininity (PF) of an input utterance using speaker recognition techniques. The estimator was designed for its clinical use and the target speakers are gender identity disorder (GID) clients, especially MtF (male to female) transsexuals. The voice therapy for MtFs is composed of three kinds of training; 1) raising the baseline F0 range, 2) changing the baseline voice quality, and 3) enhancing fo dynamics to produce an exaggerated intonation pattern. The first two focus on static acoustic properties of speech and the voice quality is mainly controlled by size and shape of the articulators, which can be acoustically characterized by the spectral envelope. Gaussian mixture models (GMM) of fo values and spectrums were built separately for biologically male speakers and female ones. Using the four models, PF was estimated automatically for each of 142 utterances of 111 MtFs. The estimated values were compared with the PF values obtained through listening tests. Results showed very high correlation (R=0.86), which is comparable to the intra-rater correlation.

ECCS2012 makes PCs and printers in computer labs accessible from off-campus environment

Information Technology Center, The University of Tokyo provides Educational Campuswide Computing System (ECCS) to approximately 60,000 users at our university. ECCS includes computer labs for class and self-study and a mail service for staff and students. ECCS2012, launched in March 2012, focuses on integrations of hardware and various channels to access the system. For example, a single kind of terminals in the labs, an integrated file server for the terminals and the mail service, and a printing service portal which connects printers in the labs to off-campus PCs and terminals in the labs to off-campus printers in commercial stores. In this paper, we describe the detail and usage statistics of ECCS2012.